"Alternative symbol" synonyms & "Formerly" #132
Conversation
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
9891e8c
to
259e0d9
Compare
- Update: Now adding synonyms for alt & included symbols, w/ type of mondo#abbreviation. - Misc updates: Todo comments. Renamed variables/methods for consistency and accuracy. Refactored some things.
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
259e0d9
to
48fc895
Compare
joeflack4
changed the title
joeflack4
changed the title
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
9ef0d72
to
8a07fa3
Compare
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
8a07fa3
to
830fbcf
Compare
- In previous commits, stated that this would also include 'included' symbols, but for now we have decided not to proceed with that. - Update: Now adding synonyms for alternative symbols, w/ type of mondo#abbreviation. - Update: Now stripping the text ', INCLUDED' from symbols. Previously was only doing for titles. - Update: For included & alternative titles/symbols ending in ', FORMERLY', these are being added as relatedSynonyms, and also being marked as owl:deprecated. - Misc updates: Updated some comments and codestyle.
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
830fbcf
to
92780ed
Compare
- Bug fix: mondo#ABBREVIATION URI - Update: Typo and codestyle of cleanup_label()
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
|
@joeflack4 do you plan to update this to use |
omim2obo/main.py
Outdated
| graph: Graph, source: URIRef, prop: URIRef, target: Union[Literal, str, URIRef], | ||
| anno_pred_vals: List[Tuple[URIRef, Union[Literal, str, URIRef]]] | ||
| ): | ||
| """Add an axion annotation to the graph.""" |
| ) -> Tuple[List[str], List[str], List[str], List[str]]: | ||
| """Separate current title/symbols from deprecated (marked 'former') ones""" | ||
| former_titles = [x for x in titles if ', FORMERLY' in x.upper()] | ||
| former_symbols = [x for x in symbols if ', FORMERLY' in x.upper()] |
There was a problem hiding this comment.
Good to apply this to symbols that are in the "Alternative Title(s); symbol(s)" column since 'FORMERLY' is used with symbols too, e.g. https://www.omim.org/entry/606417?search=606417&highlight=FORMERLY
There was a problem hiding this comment.
Currently, it looks like I only made this change to "alternative titles/symbols" actually, satisfying:
But this does not satisfy also doing this for "included" as for:
I'm really surprised I didn't do that. I thought I had!
- @joeflack4 Deal with "formerly" "included" entries
There was a problem hiding this comment.
Oh, one reason I hadn't done this is because of:
For now I'm adding them as mondo#omim_included. I don't think we can mark them as a synonym type unless we start doing that with "included titles" as well.
There was a problem hiding this comment.
Complete. Now we are processing all of the "former" entries:
- alt titles
- alt symbols
- included titles
- included symbols
| titles2 = [remove_included_and_formerly_suffixes(x) for x in titles] | ||
| symbols2 = [remove_included_and_formerly_suffixes(x) for x in symbols] | ||
| # additional reformatting for titles | ||
| titles2 = [cleanup_label(x) for x in titles2] # todo: is this redundant? gets done when adding synonyms to graph |
There was a problem hiding this comment.
try to sort out this answer and remove the #todo here
There was a problem hiding this comment.
Actually, it's not really this line that's redundant, it's several places further down in the process.
This became more involved than I anticipated. I did it as a separate PR:
Even though this line titles2 = [cleanup_label(x) for x in titles2] is new, those changes really aren't in scope for this PR, so I think we can address that PR separately and merge it into main later.
omim2obo/main.py
Outdated
| if inc_titles_str: | ||
| included_titles, included_symbols = parse_title_symbol_pairs(inc_titles_str) | ||
| included_titles, included_symbols = clean_alt_and_included_titles(included_titles, included_symbols) | ||
| included_is_included = included_titles or included_symbols # redundant. can't be included symbol w/out title |
There was a problem hiding this comment.
are you doing this so you don't have to check for an empty list to see if the entry has any values in the "Included Title(s); symbols" column?
There was a problem hiding this comment.
Naw. The reason I did this is to make the code clearer. Didn't want to come across the code just checking for "included titles" and have someone thing "well why isn't it checking for included symbols?".
Come to think of it, another approach is that I could remove this line. And instead, further down, I could potentially replace if included_is_included with something like if included_titles: # fyi: only need to check for included titles because there can't be any included symbols unless there are included titles.
There was a problem hiding this comment.
yes, it is a somewhat oddly named variable and duplicates information, included_titles, that already exists in this case an could be used in that later check
omim2obo/main.py
Outdated
| graph.add((omim_uri, RDFS['comment'], Literal(included_detected_comment))) | ||
| for included_label in cleaned_inc_labels: | ||
| graph.add((omim_uri, URIRef(INCLUDED_URI), Literal(label_cleaner.clean(included_label, abbrev)))) | ||
| # - exact titles |
There was a problem hiding this comment.
nit: it would be more clear if you maintained the omim terminology of Preferred or Alternative since there is not an "exact title"
There was a problem hiding this comment.
Ah! I would agree with you.
Actually, AFAIK I've dropped the previous "exact" nomenclature, which I think was referencing "preferred" previously, everywhere in the codebase.
What 'exact' means here is 'exact synonyms: titles'.
Below it I have 'related titles', 'exact abbreviations', 'related, deprecated 'former' titles', and 'related, deprecated 'former' abbreviations'.
If you want, I can update all these comments to say "exact/related synonyms".
There was a problem hiding this comment.
if you are not going to use the terminology from the mimTitles file, then document somewhere what "exact titles" is, for example the value from the "Preferred Title; symbol" column for the "Preferred Title" after processing and now being added as an exact synonym or 'exact synonyms: titles' or the processed title value added as an exact synonym
This kind of documentation should be done for each of the types below, e.g. exact abbreviations, etc.
It's easy to read that the information is being added as an exact synonym for example, but since "exact title" is something made up and there is a lot of processing of the value/variable (pref_title) being added here there is a fair bit of backtracking one needs to do to sort out what the original value from the file is being added here since there nothing in the source file named "exact title"
There was a problem hiding this comment.
I don't think I successfully communicated earlier. The comment isn't inspired by the previously named variables. I've since refactored all the variable names to be in sync with the OMIM data model.
The 'exact' refers to 'exact synonym' I mean to say that this is a section for "any title (non symbol) that is an exact synonym". Easier to see if you look at this line with the lines around it:
# Add synonyms
# - exact titles
graph.add((omim_uri, oboInOwl.hasExactSynonym, Literal(label_cleaner.clean(pref_title, pref_abbrev))))In any case, I have refactored these comments! They will be easier to read now and will not cause any confusion.
omim2obo/main.py
Outdated
| for title in former_alt_titles: | ||
| clean_title = Literal(label_cleaner.clean(title, pref_abbrev)) | ||
| add_triple_and_optional_annotations(graph, omim_uri, oboInOwl.hasRelatedSynonym, clean_title, | ||
| [(OWL.deprecated, Literal(True))]) |
There was a problem hiding this comment.
I do not think that the intent was to add owl:deprecated as an annotation on the synonym annotation. I have a question into Sabrina to confirm this
There was a problem hiding this comment.
Hmm, OK. Right now I can't think of anything else I would deprecate. I wouldn't want to deprecate the whole term, since non-deprecated MIMs can have both deprecated and non-deprecated synonyms.
- @twhetzel to discuss "formerly" title/symbol deprecation with Sabrina
There was a problem hiding this comment.
I don't think there is a need to add owl:deprecated to anything here
There was a problem hiding this comment.
Hmm... I'm basing this change off of:
In that issue it links to the discussion with Nicole where she talks about this being their standard practice.
Example:
synonym: "juvenile carcinoma (formerly)" RELATED DEPRECATED [GARD:0009408]
This change makes it so that this process is automated for them.
edit: FYI @twhetzel you also asked me to make this change it looks like: comment
There was a problem hiding this comment.
Ok, let's confirm with @sabrinatoro which of the suggestions to follow since the comments in that issue from Sabrina and Nicole are not in agreement.
omim2obo/main.py
Outdated
| for abbrevs in [pref_symbols, alt_symbols]: | ||
| for abbreviation in abbrevs: | ||
| add_triple_and_optional_annotations(graph, omim_uri, oboInOwl.hasExactSynonym, Literal(abbreviation), | ||
| [(OBOINOWL.hasSynonymType, MONDONS.abbreviation)]) |
There was a problem hiding this comment.
I think this needs updating from other decisions, e.g. oboInOwl and use the OMO abbreviation property
There was a problem hiding this comment.
@twhetzel For those types of situations, my solution is usually (a) if I merge those first, I merge main back into this PR, or (b) if I merge this first, then when those get merged, those changes make it into main anyhow.
Usually I opt for 'a' for best results.
But sometimes it is better to be duplicative about this stuff. Builds do run off of main after all. Actually, one just ran and released, but I deleted it because it did not include this OMO bug fix.
Anyway I'll go ahead and propagate those changes to this PR, perhaps just by merging those feature branches into it now.
Great review btw! Thanks.
There was a problem hiding this comment.
I went ahead and merged in main (after merging in the other PRs you approved) as well as:
So now those changes have propagated here.
I also added this warning to the OP:
Warning
Merging this will also merge the following PR, so it should be completed first:
- Update: Docstring to make clearer - Update: Typo fix - Update: Now "former" included titles/symbols are also being processed
- Update: Comments to improve clarity - Delete: redundant variable 'included_is_included', and refactored / recommented around that change.
- Formerly: synonym pred: related -> exact - Formerly: owl:drepecated synonym annotation -> MONDO:omim_formerly
joeflack4
force-pushed
the
abbrevs-alt-n-included
branch
from
b509332
to
112a19f
Compare
oio:exactSynonym&mondo#abbreviation#139, INCLUDED#141, FORMERLY#137oio:exactSynonym&MONDO:omim_formerly#138Google drive folder with outputs and diff.
Changes
Add alt & included symbol synonyms
'Draft' status
Joe 10/03: I believe that from my end, requested changes and all else is done. But I think Trish wanted more time to review, so to prevent accidental merge we put in draft mode.